On Wed, May 23, 2018 at 01:40:34PM +0300, Jani Nikula wrote: > On Wed, 23 May 2018, Daniel Vetter <[email protected]> wrote: > > Just checks that merges aren't done without minimal thought. > > > > Cc: Jani Nikula <[email protected]> > > Cc: Dave Airlie <[email protected]> > > Signed-off-by: Daniel Vetter <[email protected]> > > --- > > dim | 22 ++++++++++++++++++++++ > > 1 file changed, 22 insertions(+) > > > > diff --git a/dim b/dim > > index 2fa158231228..ff9e572726cb 100755 > > --- a/dim > > +++ b/dim > > @@ -785,6 +785,24 @@ function checkpatch_commit_push > > return $rv > > } > > > > +function checkmerge_commit_push > > +{ > > + local sha1 managed_branch rv body_text > > + > > + sha1=$1 > > + managed_branch=${2} > > + rv=0 > > + > > + body_text="$(git show $sha1 -s --format="format:%b" | grep -v "^$" | > > grep -v "^\S*:")" > > + > > + if [[ -z "$body_text" ]] ; then > > + echoerr "$sha1 is lacking merge commit justification" > > + rv=1 > > + fi > > + > > + return $rv > > +} > > + > > function checkpatch_commit_push_range > > { > > local rv managed_branch > > @@ -797,6 +815,10 @@ function checkpatch_commit_push_range > > checkpatch_commit_push $sha1 $managed_branch || rv=1 > > done > > > > + for sha1 in $(git rev-list "$@" --merges) ; do > > Will that include merges within merges? If there's a backmerge of > drm-next which has a backmerge of upstream, there could be > anything... and *tons* of it.
Yes and no. For pushing we include --committer=$self in the argument list, not just the sha1 rangeish (Aside: that's why we need those last, since there's a variable amount of them). When pulling, we check the entire pull. Same applies for checking normal commits, which I think is the right thing to do for these sanity checks. More concrete: When Dave pushes, dim will only complain about his own merges. But when he pulls, dim will complain e.g. both about dinq and about gvt merges. -Daniel > > BR, > Jani. > > > + checkmerge_commit_push $sha1 $managed_branch || rv=1 > > + done > > + > > if [ $rv == "1" ] ; then > > warn_or_fail "issues in commits detected" > > fi > > -- > Jani Nikula, Intel Open Source Graphics Center -- Daniel Vetter Software Engineer, Intel Corporation http://blog.ffwll.ch _______________________________________________ dim-tools mailing list [email protected] https://lists.freedesktop.org/mailman/listinfo/dim-tools
